摘要
鉴于有些模型间不好相互比较,有些非线性模型不好检验,作者提出任何模型都可算出实测值与模型值的线性相关系数,并定名为拟合相关系数,以此作为任何模型拟合精度的评价和模型间精度比较的尺度,建议任何模型都采用对拟合相关系数进行统计检验。最后作者认为建立数学模型不能只偏爱于拟合精度的高低,还应关注模型应用性能的强弱。
Some mathematics models are not easy to be compared with each other,and some linear models are not easy to be examined. The authors put forward a related coefficient that can calculate the measured value and predicted value of any model, which is called matched related coefficient, and suggest that it be adopted to any model to carry out statistic examination. Finally, the authors believe that we should not be partial to the precision of the matched related coefficient, we should also pay more attention to the applicability of models.
出处
《江西农业大学学报》
CAS
CSCD
北大核心
2006年第4期641-644,共4页
Acta Agriculturae Universitatis Jiangxiensis
关键词
数学模型
评价
检验
mathematics model
evaluation
examination
